Association for Scientic Computing Electronics and Engineering (ASCEE): Open Journal Systems
Not a member yet
785 research outputs found
Sort by
Deep learning-based cervical lesion segmentation in colposcopic images
Get PDFArtificial intelligence assisted cancer detection has changed the ream of diagnosis precision. This study aims to propose a segmentation network using artificial intelligence for accurately segmenting the cervix region and acetowhite lesions in cervigram images, addressing the shortage of skilled colposcopists and streamlining the training process. A computational approach is employed to develop and train a deep learning model specifically tailored for cervix region and acetowhite lesion segmentation in cervigram images. A dataset acquired in collaboration with KIDWAI memorial cancer research institute is used for building the model. Cervigram images are collected for training and validation, and a deep learning architecture is constructed and trained using annotated datasets. The segmentation network based on efficientnet architecture and atrous spatial pyramid pooling is designed to accurately identify and delineate the target regions, with performance evaluation conducted using precision, accuracy, recall, dice score, and specificity metrics. The proposed segmentation network achieves a precision of 0.7387±0.1541, accuracy of 0.9291, recall of 0.7912±0.1439, dice score of 0.7431±0.1506, and specificity of 0.9589±0.0131, indicating its reliability and robustness in segmenting cervix regions and acetowhite lesions in cervigram images. This research demonstrates the feasibility and effectiveness of using artificial intelligence-based computational models for cervix region and acetowhite lesion segmentation in cervigram images. It provides a foundation for further investigations into classifying cervix malignancy using AI techniques, potentially enhancing early detection and treatment of cervical cancer while addressing the shortage of skilled professionals in the fieldÂ
Betta fish classification using transfer learning and fine-tuning of CNN models
Get PDFBetta fish, known as freshwater fighters, are in demand because of their beauty and characteristics. These betta fish such as Crowntail, Halfmoon, Doubletail, Spadetail, Plakat, Veiltail, Paradise, and Rosetail are hard to recognize without knowledge about them. Therefore, transfer learning of Convolutional Neural Network models was proposed to classify the betta fish from the image. The transfer learning process used a pre-trained model from ImageNet of VGG16, MobileNet, and InceptionV3 and fine-tuned the models on the betta fish dataset. The models were trained on 461 images, validated with 154 images, and tested on 156 images. The result shows that the InceptionV3 model excels with 0.94 accuracies compared to VGG16 and MobileNet which acquire 0.93 and 0.92 accuracy respectively. With good accuracy, the trained model can be used in betta fish recognition applications to help people easily identify betta fish from the image
Research on Indoor 3D Reconstruction Technology Based on Semantic Visual Simultaneous Localization and Mapping
Get PDFIn response to the challenge that traditional visual simultaneous localization and mapping (SLAM) systems, based on the assumption of a static environment, struggle to achieve real-time indoor 3D reconstruction in complex dynamic scenes, this paper proposes a real-time indoor 3D reconstruction algorithm based on semantic visual SLAM. By leveraging object detection to obtain 2D semantic information and providing prior information for geometric methods, the fusion of the two effectively suppresses dynamic features, reduces reliance on deep learning methods, and ensures the algorithm's real-time performance. Experimental results on dynamic scenes in the TUM RGB-D dataset show that our algorithm maintains nearly unchanged real-time performance while achieving an average performance improvement of approximately 97.56% and 97.31% on the TUM dataset and Bonn dataset, respectively, compared to the ORB-SLAM2 system. Moreover, our algorithm can reconstruct more intuitive indoor global Octo-map and semantic metric maps compared to sparse point cloud maps, effectively enhancing the scene perception capability of mobile robots and laying the foundation for performing advanced tasks. Furthermore, our algorithm demonstrates a 3.5-10.5 times improvement in real-time performance compared to other mainstream semantic SLAM systems. Experimental results on the NVIDIA Jetson AGX Xavier confirm that our algorithm can run in real time on low-power platforms such as mobile robots or drones. However, the drawbacks of our algorithm include lower reconstruction accuracy in low-texture and large-scale scenes and ineffective suppression of dynamic features in low-dynamic scenes. Future work will consider replacing and improving deep learning methods and integrating IMU and other sensors to enhance system usability
Decision tree based algorithms for Indonesian Language Sign System (SIBI) recognition
Get PDFIndonesian Sign Language System (SIBI) recognition plays a crucial role in improving effective communication for individuals with hearing loss in Indonesia. To support automatic SIBI recognition, this research presents a performance analysis of two main algorithms, namely Decision Tree and C4.5, in the context of the SIBI recognition task. This research utilizes a rich SIBI dataset that includes a variety of SIBI signs used in everyday communication. Data pre-processing, model construction with both algorithms, and model performance evaluation using accuracy, precision, recall, and F1-score metrics are all part of the study. Regarding SIBI recognition accuracy, the experimental results demonstrate that the Decision Tree performs better than Decision Tree. The Decision Tree also makes models that are easier to understand, which is important for making communication systems based on SIBI
Gender conflict resolution in Nigerian and African American standup performances
Get PDFStand-up comedy, a prominent facet of live theatrical entertainment, occupies a unique position in the entertainment industry, embodying a key element of popular culture. This comedic form reflects the sociological dynamics of both historical and contemporary societies, influencing and engaging both local and international audiences. Beyond providing psycho-physical therapy for performers, stand-up comedy plays a pivotal role in shaping societal perceptions. This article focuses on a gender-focused analysis of stand-up comedy, exploring its implications for domestic growth, marriage unity, and love. Utilizing three digital recordings featuring stand-up comedians, namely Bright Okpocha (Basket Mouth), Ayo Makun (AY), and Eddie Murphy, the study examines their unique approaches to addressing gender-based violence and resolving conflicts. The selection criteria were based on the comedians' distinct counseling methods and perspectives on gender-related issues. Drawing on Richard Schechner's performance theory, the analysis employs both performance and content analyses to elucidate the spectrum of gender conflicts and resolutions presented by the comedians. The findings underscore the importance of mutual understanding and communication in resolving gender conflicts. The article also delves into the influence of parental roles in household management and the pursuit of material wealth. Emphasizing the significant role of counseling modes, the study concludes that stand-up comedy, with its oratory counseling mode and therapeutic elements, serves as an effective means of addressing gender conflicts, contributing to mutual understanding and successful resolutions within society
Mapping Art: 3D geo-visualization and virtual worlds in cultural heritage
Get PDFThe purpose of this paper is to highlight the use of digital mapping and GIS in combination with 3D geo-visualization in depicting landscapes and cultural heritage sites. As computers evolve in their processing capabilities (graphics card, processor, memory speed), so does the graphical output representing our world that can be done in less time. Short processing times and available highly sophisticated software create conditions to test existing models of our 3D world or best create alternative renderings of monuments and landscapes. A study area has been chosen to present such a methodology. Ancient Pylos was selected as the study area due to the abundance of bibliographic references available, facilitating the creation of a cartographic representation of the region based on archaeological discoveries and historical sources. The advantages of employing 3D geo-visualizations in archaeology are numerous and are examined within this article. Visual imaging aids archaeologists in elucidating intricate or deficient details pertaining to a monument, simultaneously streamlining archaeological data for easier comprehension by the general public. The use of 3D geo-visualization to represent data and other non-photorealistic details is expected to dominate in the near future
Wireless Sensor Networks Fault Detection and Identification
Get PDFWe have developed and experimentally tested a set of models for the detection and identification of sensor faults that commonly occur in wireless sensor networks. Considered faults include outlier, spike, variance, high-frequency noise, offset, gain, and drift faults. These faults affect the system operations and can endanger operators, final users, and the general public. The fault detection models are divided into two classes: data-centric models, which only analyze a single data stream, and system-centric models, which consider the overall system. For data-centric models, we use the magnitude, the gradient, and the variance of raw sensor data to model faults. For system-centric models, we introduce variogram-based techniques that allow faults to be detected by comparing readings from multiple sensors that measure related phenomena. For data-centric and system-centric sensor fault detection, we show how a few model parameters affect the sensitivity of wireless sensor network fault models. We present simulation and experimental results that illustrate the fault detection and identification models. The system is intended for health monitoring applications of the NASA Stennis Space Center (SSC) test stands and widely distributed support systems, including pressurized gas lines, propellant delivery systems, and water coolant lines. The testbed consists of Coremicro® reconfigurable embedded smart sensor nodes [29] capable of wireless communication, a network-capable application processor, a wireless base station, the software that supports sensor and actuator health monitoring, a database server, and a smartphone running a health monitoring Android application
Integrating ChatGPT into EFL writing instruction: Benefits and challenges
Get PDFTeaching English as a Foreign Language (EFL) writing skills presents a unique set of challenges, from developing students' language proficiency to providing personalized feedback and support. In recent years, there has been growing interest in the potential of AI and NLP, such as ChatGPT, to support language education. This paper explores the potential benefits and challenges of using ChatGPT in EFL writing instruction. This paper is based on a review of relevant literature and highlights the potential of ChatGPT for EFL writing students. ChatGPT's ability to generate human-like text based on the input it receives makes it a potentially helpful tool for supporting students in their writing practice. Its natural language capabilities can engage and motivate learners. However, the potential challenges of using ChatGPT in language education are also discussed, such as the design, implementation and potential ethical concerns. This paper aims to provide a comprehensive overview of the potential benefits and challenges of integrating ChatGPT into EFL writing instruction. By exploring the potential of ChatGPT as a tool for language education, we aim to contribute to the ongoing debate on the role of technology and provide guidance for educators interested in using ChatGPT in their classrooms
YOLOv3 and YOLOv5-based automated facial mask detection and recognition systems to prevent COVID-19 outbreaks
Get PDFObject detection system in light of deep learning have been monstrously effective in complex item identification task images and have shown likely in an extensive variety of genuine applications counting the Coronavirus pandemic. Ensuring and enforcing the proper use of face masks is one of the main obstacles in containing and reducing the spread of the infection among the population. This paper aims to find out how the urban population of a megacity uses facial masks correctly. Using YOLOv3 and YOLOv5, we trained and validated a brand-new dataset to identify images as "with mask", "without mask", and "mask not in position". In the YOLOv3 we carried out three pre-trained models which are: YOLOv3, YOLOv3-tiny, and SPP-YOLOv3. In addition, we utilized five pre-trained models in the YOLOv5: YOLOv5n, YOLOv5s, YOLOv5m, YOLOv5l, and YOLOv5x. The dataset is included 6550 pictures with three classes. On mAP, the dataset achieved a commendable 95% performance accuracy. This research can be used to monitor the proper use of face masks in various public spaces through automated scanning
Placement model for students into appropriate academic class using machine learning
Get PDFChoosing the right academic major for junior secondary students into senior secondary school will assist both students and their teachers toward achieving the academic goal. Traditionally, students seeking admission into senior classes (Gambia, Sierra-leone, Ghana, Liberia and Nigeria) must have passed stipulated examinations like Basic Education Certificate Examination (BECE) and/or West Africa Junior Certificate Examination, which are done at the end of year three (at a sitting). They must pass the exam(s) satisfactorily with no emphasis on any of Science, Art or Commercial related subjects. Some schools use “Mock exam†or “Placement exam†as the basis for their placement of students but all are done at a sitting (end of year three). Though this method is to an extent valid but associated with some challenges (bias) as it does not carry along the student’s academic history in making decision for placement into appropriate class. However, we proposed a model that predicts appropriate academic class of Science, Art or Commercial for Junior students based on their progressive academic performances (history) of their predecessors on related subjects using ten supervised machine learning techniques. Two evaluation techniques were applied (70/30 splitting and 10-fold cross validation). The highest results of this research showed accuracy of 93% with Random forest, 98% precision with random forest, 99% recall with Decision tree and 94% f1 score with Random forest and KNN (cross validation). The correlation coefficient of the proposed model recorded 0.3 higher than that of the existing method. This research will benefit all stakeholders in education and students in particular because their academic performances over time stands a better chance for appropriate placement